A 3D-1D substitution matrix for protein fold recognition that includes predicted secondary structure of the sequence

Structure prediction in a post-genomic environment: a secondary and tertiary structural model for the initiation factor 5A family

Two predictions have been prepared for the fold of initiation factor 5A (IF5A) starting from a set of homologous sequences. In the first, a secondary structural model was predicted for the protein in 1994, when only eleven homologs (and no eubacterial homologs) had been sequenced. The second was made recently, after genome projects had generated a total of 33 sequences for the protein family from species of all three kingdoms of life. With the second set of sequences, but not with the first, it was possible to predict that the N-terminal domain of the protein folds in a possibly open beta-barrel/sandwich core structure, with a short helix capping one side of the barrel. We place the pair of predictions in the public domain before an experimental structure is known. This example illustrates the impact of genome sequencing projects on structure prediction from sequence alignments.

Sequence profile of the parallel beta helix in the pectate lyase superfamily

The parallel beta helix structure found in the pectate lyase superfamily has been analyzed in detail. A comparative analysis of known structures has revealed a unique sequence profile, with a strong positional preference for specific amino acids oriented toward the interior of the parallel beta helix. Using the unique sequence profile, search patterns have been constructed and applied to the sequence databases to identify a subset of proteins that are likely to fold into the parallel beta helix. Of the 19 families identified, 39% are known to be carbohydrate-binding proteins, and 50% belong to a broad category of proteins with sequences containing leucine-rich repeats (LRRs). The most striking result is the sequence match between the search pattern and four contiguous segments of internalin A, a surface protein from the bacterial pathogen Listeria monocytogenes. A plausible model of the repetitive LRR sequences of internalin A has been constructed and favorable 3D-1D profile scores have been calculated. Moreover, spectroscopic features characteristic of the parallel beta helix topology in the pectate lyases are present in the circular dichroic spectrum of internalin A. Altogether, the data support the hypothesis that sequence search patterns can be used to identify proteins, including a subset of LRR proteins, that are likely to fold into the parallel beta helix.

Random mutagenesis of 1-aminocyclopropane-1-carboxylate synthase: a key enzyme in ethylene biosynthesis

1-Aminocyclopropane-1-carboxylate synthase (ACC synthase, EC 4.4.1. 14) catalyzes the rate-limiting step in the ethylene biosynthetic pathway in plants. To determine the amino acid residues critical for the structure and function of this enzyme, the tomato Le-ACS2 isoenzyme has been subjected to both site-directed and PCR random mutagenesis. Mutant ACC synthases with reduced enzyme activity have been selected by using a genetic screen based on the functional complementation of an Escherichia coli Ile auxotroph that has been engineered to express ACC deaminase from Pseudomonas sp. The DNA sequence of almost 1,000 clones has been determined, and 334 single missense mutations have been selected for analysis. We have identified three classes of mutants based on their activity and expression in E. coli. Class I and II mutants have the same level of protein expression as the wild type, but their enzyme activity is reduced to 0-5% and 5-50%, respectively. Class III mutants have neither activity nor detectable protein expression. The inactive mutations are clustered in regions that are highly conserved among various ACC synthases. This library of mutants will facilitate the elucidation of structure-function relationships of this regulatory enzyme.

Deciphering cryptic similarities in protein binding sites

Rapid expansion in the number of plausible drug targets arising from genomics research has created new pressures for increased efficiency in discovery of high specificity candidate drug compounds. Improved understanding of conserved features among protein structures provides a promising route to achieving this goal. Indirect evidence implies that important similarities are now ripe for elucidation by emerging experimental approaches.

Protein structure prediction

Genome sequencing projects continue to provide a flood of new protein sequences, and prediction methods remain an important means of adding structural information. Recently, there have been advances in secondary structure prediction, which feed, in turn, into improved fold recognition algorithms. Finally, there have been technical improvements in comparative modelling, and studies of the expected accuracy of three-dimensional structural models built by this method.

First partial three-dimensional model of human monoamine oxidase A

A survey of the major known structural aspects of monoamine oxidase (MAO) is given and a first partial model of human MAO A is presented. This 3D model has been established using secondary structure predictions and fold recognition methods. It shows two alpha/beta domains (the FAD-binding N-terminal and central domains) and an alpha+beta domain. The C-terminal region is predicted to be responsible for anchoring the protein into the mitochondrial membrane and was not modeled. The covalent binding of the flavin cofactor to a cysteine residue is well predicted. The model is validated with experimental data from the literature and should be useful in designing new experimental studies (site-directed mutagenesis, chemical modification, specific antibodies). This first step towards the 3D structure of monoamine oxidase should contribute to a better understanding of the mechanisms of action and inhibition of this drug target in the treatment of clinical depression.

Unification of protein families

Computational biology exploits the evolutionary connectivity between proteins and protein families to predict structural and functional properties of uncharacterized gene products. In the past year, conceptual and statistical refinements have substantially improved algorithms for the detection of remote homologues. In conjunction with the rapid growth of biological databases, the global organization of proteins into sequence families, functional families and structural families has become both pertinent and feasible.

Fold prediction by a hierarchy of sequence, threading, and modeling methods

Several fold recognition algorithms are compared to each other in terms of prediction accuracy and significance. It is shown that on standard benchmarks, hybrid methods, which combine scoring based on sequence-sequence and sequence-structure matching, surpass both sequence and threading methods in the number of accurate predictions. However, the sequence similarity contributes most to the prediction accuracy. This strongly argues that most examples of apparently nonhomologous proteins with similar folds are actually related by evolution. While disappointing from the perspective of the fundamental understanding of protein folding, this adds a new significance to fold recognition methods as a possible first step in function prediction. Despite hybrid methods being more accurate at fold prediction than either the sequence or threading methods, each of the methods is correct in some cases where others have failed. This partly reflects a different perspective on sequence/structure relationship embedded in various methods. To combine predictions from different methods, estimates of significance of predictions are made for all methods. With the help of such estimates, it is possible to develop a "jury" method, which has accuracy higher than any of the single methods. Finally, building full three-dimensional models for all top predictions helps to eliminate possible false positives where alignments, which are optimal in the one-dimensional sequences, lead to unsolvable sterical conflicts for the full three-dimensional models.

Fold assembly of small proteins using monte carlo simulations driven by restraints derived from multiple sequence alignments

The feasibility of predicting the global fold of small proteins by incorporating predicted secondary and tertiary restraints into ab initio folding simulations has been demonstrated on a test set comprised of 20 non-homologous proteins, of which one was a blind prediction of target 42 in the recent CASP2 contest. These proteins contain from 37 to 100 residues and represent all secondary structural classes and a representative variety of global topologies. Secondary structure restraints are provided by the PHD secondary structure prediction algorithm that incorporates multiple sequence information. Predicted tertiary restraints are derived from multiple sequence alignments via a two-step process. First, seed side-chain contacts are identified from correlated mutation analysis, and then a threading-based algorithm is used to expand the number of these seed contacts. A lattice-based reduced protein model and a folding algorithm designed to incorporate these predicted restraints is described. Depending upon fold complexity, it is possible to assemble native-like topologies whose coordinate root-mean-square deviation from native is between 3.0 A and 6.5 A. The requisite level of accuracy in side-chain contact map prediction can be roughly 25% on average, provided that about 60% of the contact predictions are correct within +/-1 residue and 95% of the predictions are correct within +/-4 residues. Precision in tertiary contact prediction is more critical than absolute accuracy. Furthermore, only a subset of the tertiary contacts, on the order of 25% of the total, is sufficient for successful topology assembly. Overall, this study suggests that the use of restraints derived from multiple sequence alignments combined with a fold assembly algorithm holds considerable promise for the prediction of the global topology of small proteins.

Seeking an ancient enzyme in Methanococcus jannaschii using ORF, a program based on predicted secondary structure comparisons

We have developed a simple procedure to identify protein homologs in genomic databases. The program, called ORF, is based on comparisons of predicted secondary structure. Protein structure is far better conserved than amino acid sequence, and structure-based methods have been effective in exploiting this fact to find homologs, even among proteins with scant sequence identity. ORF is a secondary structure-based method that operates solely on predictions from sequence and requires no experimentally determined information about the structure. The approach is illustrated by an example: Thymidylate synthase, a highly conserved enzyme essential to thymidine biosynthesis in both prokaryotes and eukaryotes, is thought to be used by Archaea, but a corresponding gene has yet to be identified. Here, a candidate thymidylate synthase is identified as a previously unassigned open reading frame from the genome of Methanococcus jannaschii, viz., MJ0757. Using primary structure information alone, the optimally aligned sequence identity between MJ0757 and Escherichia coli thymidylate synthase is 7%, well below the threshold of sensitivity for detection by sequence-based methods.

Histone acetyltransferase activity of yeast Gcn5p is required for the activation of target genes in vivo

Gcn5p is a transcriptional coactivator required for correct expression of various genes in yeast. Several transcriptional regulators, including Gcn5p, possess intrinsic histone acetyltransferase (HAT) activity in vitro. However, whether the HAT activity of any of these proteins is required for gene activation remains unclear. Here, we demonstrate that the HAT activity of Gcn5p is critical for transcriptional activation of target genes in vivo. Core histones are hyperacetylated in cells overproducing functional Gcn5p, and promoters of Gcn5p-regulated genes are associated with hyperacetylated histones upon activation by low-copy Gcn5p. Point mutations within the Gcn5p catalytic domain abolish both promoter-directed histone acetylation and Gcn5p-mediated transcriptional activation. These data provide the first in vivo evidence that promoter-specific histone acetylation, catalyzed by functional Gcn5p, plays a critical role in gene activation.

Modeling the paramyxovirus hemagglutinin-neuraminidase protein

The paramyxovirus hemagglutinin-neuraminidase (HN) protein exhibits neuraminidase activity and has an active site functionally similar to that in influenza neuraminidases. Earlier work identified conserved amino acids among HN sequences and proposed similarity between HN and influenza neuraminidase sequences. In this work we identify the three-dimensional fold and develop a more detailed model for the HN protein, in the process we examine a variety of protein structure prediction methods. We use the known structures of viral and bacterial neuraminidases as controls in testing the success of protein structure prediction and modeling methods, including knowledge-based threading, discrete three-dimensional environmental profiles, hidden Markov models, neural network secondary structure prediction, pattern matching, and hydropathy plots. The results from threading show that the HN protein sequence has a 6 beta-sheet propellor fold and enable us to assign the locations of the individual beta-strands. The three-dimensional environmental profile and hidden Markov model methods were not successful in this work. The model developed in this work helps to understand better the biological function of the HN protein and design inhibitors of the enzyme and serves as an assessment of some protein structure prediction methods, especially after the x-ray crystallographic solution of its structure.

Assigning folds to the proteins encoded by the genome of Mycoplasma genitalium

A crucial step in exploiting the information inherent in genome sequences is to assign to each protein sequence its three-dimensional fold and biological function. Here we describe fold assignment for the proteins encoded by the small genome of Mycoplasma genitalium. The assignment was carried out by our computer server (http://www.doe-mbi.ucla.edu/people/frsvr/ frsvr. html), which assigns folds to amino acid sequences by comparing sequence-derived predictions with known structures. Of the total of 468 protein ORFs, 103 (22%) can be assigned a known protein fold with high confidence, as cross-validated with tests on known structures. Of these sequences, 75 (16%) show enough sequence similarity to proteins of known structure that they can also be detected by traditional sequence-sequence comparison methods. That is, the difference of 28 sequences (6%) are assignable by the sequence-structure method of the server but not by current sequence-sequence methods. Of the remaining 78% of sequences in the genome, 18% belong to membrane proteins and the remaining 60% cannot be assigned either because these sequences correspond to no presently known fold or because of insensitivity of the method. At the current rate of determination of new folds by x-ray and NMR methods, extrapolation suggests that folds will be assigned to most soluble proteins in the next decade.